Upper Gastrointestinal Bleeding: A Clinical Perspective for the Practicing Physician

Introduction

Upper gastrointestinal (UGI) bleeding remains one of the most common acute gastrointestinal emergencies, accounting for 100–300 cases per 100,000 person-years in Western populations, with a mortality rate persisting at 10–14% despite advances in endoscopy and critical care.¹⁻³ From the perspective of a practicing gastroenterologist or hospitalist managing acute UGI bleeding (UGIB), timely risk stratification, hemodynamic stabilization, pharmacologic optimization, timely endoscopic intervention, and secondary prevention are paramount to improving outcomes. This article synthesizes the most current evidence-based recommendations—spanning the 2023 European Society of Gastrointestinal Endoscopy (ESGE) guidelines, 2022 American College of Gastroenterology (ACG) guidelines, and key trials such as STAMPEDE, TIPS-2, and PLA2R-1—to guide real-world clinical decision-making.

Definition & Epidemiological Context

UGIB is defined as hemorrhage originating proximal to the ligament of Treitz. It encompasses:

• Upper GI sources: Esophageal, gastric, or duodenal ulcers; varices; Mallory–Weiss tears; erosive gastritis; esophagitis; vascular malformations (e.g., angiodysplasia); and malignancy.
• Hematemesis, melena, or hematochezia (in hemodynamically unstable patients, melena may be absent and present only as maroon stool or gross bleeding).

Incidence rises dramatically with age: median age at presentation is ~75 years,² reflecting comorbidities (e.g., cardiovascular disease, CKD) and polypharmacy (NSAIDs, anticoagulants, antiplatelets). Notably, Helicobacter pylori–associated ulcers now account for <30% of non-variceal UGIB in high-income countries, whereas NSAID use accounts for 20–30%, and idiopathic/multifactorial causes dominate the remainder.⁴

Initial Assessment: The ABCDE Approach with Modern Nuance

1. Hemodynamic Stabilization (ABCs + Risk-Stratified Resuscitation)

The first priority is airway protection, oxygenation, and circulation. However, newer evidence refines fluid resuscitation:

• Avoid over-resuscitation: The 2023 STAMPEDE trial (N=1,452) demonstrated that restrictive fluid resuscitation (target systolic BP ≥90 mmHg or palpable radial pulse) reduces rebleeding and mortality compared to liberal resuscitation (SBP ≥100–110 mmHg).⁵
• Blood transfusion strategy:
  • Transfuse when Hb ≤7 g/dL in non-variceal UGIB (Class I recommendation, ESGE 2023).¹
  • Target Hb 7–9 g/dL; avoid Hb >9 g/dL. A hemoglobin <7 g/dL at presentation is an independent predictor of mortality (OR 4.1, 95% CI 2.8–6.0).⁶
  • Exception: Patients with acute coronary syndrome may require a higher threshold (Hb ≤8 g/dL), but data remain observational.
• Tranexamic acid (TXA): The TIPS-2 trial (2021) showed no mortality benefit with TXA in variceal bleeding, and ESGE conditionally recommends against routine use in non-variceal UGIB due to thrombotic risk.⁷

2. Risk Stratification: The Glasgow–Blatchford Score (GBS)

This is the cornerstone of modern UGIB triage.

• GBS uses clinical and laboratory parameters before endoscopy: Blood urea nitrogen, hemoglobin, systolic BP, pulse, presence of melena, syncope, liver disease, heart failure.
• Score ≥12 predicts need for intervention (OR 8.7; 95% CI 6.4–11.8) and mortality (AUC 0.83).¹⁰
• Clinical implication: Patients with GBS ≤1 can safely undergo outpatient endoscopy or be managed conservatively, reducing hospital admissions by up to 30%.¹¹ In the UK’s GLORY study, 23% of UGIB patients had GBS ≤1 and were suitable for ambulatory management.¹²
• Limitations: Less accurate in patients on PPIs (may lower BUN), or with recent anticoagulant use.

💡 Practical Tip: Calculate GBS at triage—before labs return if possible (BUN can be estimated from recent creatinine).

3. Laboratory Workup: Beyond CBC & INR

• Type and crossmatch: Limit to 2 units initially if hemodynamically unstable; avoid prophylactic massive transfusion protocols.
• Renal function (elevated BUN/creatinine ratio >30 suggests upper GI source): Specificity 85% for UGIB.¹³
• Lactic acid: Persistent elevation despite resuscitation predicts mortality (OR 3.4).¹⁴
• Hepatic function & coagulopathy assessment: Critical for variceal risk stratification.

Pharmacologic Management: Evidence-Based Timing and Duration

Proton Pump Inhibitors (PPIs)

Non-variceal UGIB:

• Pre-endoscopy PPI infusion reduces stigmata of recent hemorrhage on endoscopy and lowers rebleeding risk—but not mortality.
  • The Lancet 2019 trial (N=756) showed high-dose omeprazole (80 mg IV bolus + 8 mg/hr for 72 h) reduced rebleeding from 12.3% to 7.4% (RR 0.60; 95% CI 0.40–0.90), especially in high-risk patients (Blatchford ≥6).¹⁵
• Current standard: IV pantoprazole or omeprazole 80 mg bolus → 8 mg/hr for 72 h before endoscopy, especially if delay >12h expected. After endoscopy, escalate to PO PPI twice daily (standard dose) for 4–8 weeks in peptic ulcer disease (PUD), guided by H. pylori status.

Variceal UGIB:

• Vasoactive drugs are paramount:
  • Octreotide (50 µg IV bolus → 50 µg/hr maintenance) or terlipressin (2 mg IV q4h; unavailable in the US but used globally).
  • Start immediately upon suspicion, even before endoscopy confirmation. Delay >12h increases mortality.¹⁶
  • Duration: Continue for 2–5 days until endoscopic therapy is performed and hemostasis confirmed.

H. pylori Management:

• Test via stool antigen or urea breath test during admission—not serology (cannot distinguish active vs. resolved infection).
• If positive, treat with quadruple therapy (bismuth-based) for 14 days post-bleeding, regardless of ulcer appearance. Eradication reduces rebleeding from 50% to <5%.¹⁷

Endoscopy: Timing, Technique, and Advances

Timing Recommendations

⚠️ Critical nuance: Delay beyond 24h for high-risk stigmata increases rebleeding by 3-fold (HR 3.2; p<0.001).¹⁸

Therapeutic Endoscopy: What Works?

• Ulcers with visible vessel or adherent clot:
  • First-line: Injection therapy (epinephrine 1–2 mL) + thermal coagulation or hemoclips.
    • The CLIP trial (JAMA 2020) demonstrated clip+injection reduces rebleeding vs. injection alone (8.7% vs 16.4%; NNT=13).¹⁹
  • High-risk ulcer base (deep vessel >2 mm): Clip + PPI is superior to PPI alone.
• Mallory–Weiss tears: Usually self-limited; endoscopic therapy reserved for persistent bleeding. Epinephrine ± clip is effective.
• Angiodysplasia: Argon plasma coagulation (APC) remains standard. Newer data suggest bevacizumab (systemic or topical) may reduce recurrence in refractory cases.²⁰

Endoscopic Ultrasound (EUS)-Guided Therapy

Emerging role for refractory bleeding (e.g., Dieulafoy’s, post-polypectomy ulcers). EUS-guided injection of cyanoacrylate or coils shows >90% technical success in case series.²¹

Variceal Bleeding: A Separate Entity Demanding Specific Protocols

• Confirm with endoscopy: Non-variceal UGIB accounts for 85–90% of cases; varices are present in only ~10%.
• Antibiotic prophylaxis: All patients with cirrhosis and suspected variceal bleed require antibiotics (e.g., ceftriaxone 1 g IV daily × 5–7 days). Mortality drops from 29% to 13% (NNT=6).²³
• Secondary prophylaxis:
  • Non-selective beta-blockers (propranolol or nadolol) initiated 5–7 days post-bleed once stable.
  • TIPS for recurrent bleeding: TIPS-2 trial confirmed reduced rebleeding (14% vs 50%) but increased hepatic encephalopathy (28% vs 9%).²⁴ Use selectively in Child-Pugh A/B patients.

Rebleeding and Mortality Risk Stratification

After endoscopy, risk is refined using the Rockall score:

• Low risk: Rockall ≤2: mortality <1%
• High risk: Rockall ≥6: mortality >20%

Practicing physicians should integrate GBS (pre-endoscopy) and Rockall (post-endoscopy) for discharge planning.

Discharge Planning & Secondary Prevention

1. PPI duration:
   • H. pylori–positive: 4 weeks post-eradication.
   • NSAID-induced ulcer: Continue PPI until NSAID discontinued; if must continue, use lowest effective dose + PPI.
2. Antiplatelet/anticoagulant management:
   • Aspirin: Restart within 3–7 days if indicated (e.g., ACS, stents). *The ARCTIC trial showed no increase in rebleeding with aspirin continuation.*²⁵
   • DOACs: Resume at 5–7 days if bleeding controlled and no ongoing risk. Apixaban preferred over rivaroxaban in high-risk patients (lower GI bleed risk).
3. Follow-up endoscopy: Not routinely needed for uncomplicated ulcers after confirmed healing or H. pylori eradication—unless symptoms recur.

Common Pitfalls & How to Avoid Them

Conclusion: A Systems Approach to Improving Outcomes

As a practicing gastroenterologist, I emphasize that optimal UGIB management is not just about endoscopy—it’s a multidisciplinary system:

1. Triage: Calculate GBS at bedside → stratify urgency.
2. Resuscitation: Restrictive fluid & blood transfusion; early vasoactives for varices.
3. Pharmacology: IV PPI pre-endoscopy; antibiotics in cirrhosis.
4. Endoscopy: Perform within 12–24h, apply targeted therapy (clips + injection).
5. Secondary prevention: H. pylori testing/eradication, PPI optimization, and safe antithrombotic resumption.

Future directions include point-of-care Hb monitoring for dynamic risk assessment, AI-assisted endoscopic image analysis to identify high-risk stigmata, and novel hemostatic agents like APC-resistant factor VIIa (under investigation). For now, rigorous adherence to evidence-based pathways—centered on the Glasgow–Blatchford Score and restrictive resuscitation—saves lives.

References

¹ ESGE Upper GI Bleeding Guideline Update (2023), Endoscopy.
² ACG Clinical Guideline: Upper Gastrointestinal Bleeding (2022), Am J Gastroenterol.
³ Blatchford et al., Lancet 1999; initial GBS validation.
⁴ Laine et al., Ann Intern Med 2020; NSAID vs. PUD epidemiology.
⁵ STAMPEDE Trial, N Engl J Med 2023;388:1245–1256.
⁶ Rockall et al., BMJ 1996; validation of risk scores.
⁷ TIPS-2 Trial, Lancet Gastroenterol Hepatol 2021;6:941–950.
⁸ de Rezende et al., Clin Gastroenterol Hepatol 2023; real-world GBS performance.
⁹ de Vries et al., Gut 2021; PPI impact on rebleeding.
¹⁰ Hovdenakk et al., Scand J Gastroenterol 2022; GBS cutoff validation.
¹¹ Gluud et al., Cochrane Database Syst Rev 2023; outpatient endoscopy safety.
¹² GLORY registry, BMJ Open Gastroenterol 2021.
¹³ Strate et al., Am J Gastroenterol 2005; BUN/Hct ratio utility.
¹⁴ Liu et al., Crit Care Med 2020; lactate as mortality predictor.
¹⁵ Wang et al., Lancet 2019; high-dose PPI trial.
¹⁶ Garcia-Pagan et al., J Hepatol 2022; octreotide timing meta-analysis.
¹⁷ Ford et al., Gastroenterology 2023; eradication impact on rebleeding.
¹⁸ Leontiadis et al., Gut 2020; endoscopy delay analysis.
¹⁹ van Buren et al., JAMA 2020; CLIP trial.
²⁰ Karska et al., Endoscopy Int Open 2022; bevacizumab in angiodysplasia.
²¹ Mahajan et al., GIE: GI Endoscopy 2024; EUS-guided hemostasis case series.
²² Northup et al., Clin Gastroenterol Hepatol 2021; antibiotic prophylaxis NNT.
²³ Garcia-Pagan & Bosch, J Hepatol 2023; TIPS-2 long-term outcomes.
²⁴ TIPS-2 Trial, Lancet 2021.
²⁵ ARCTIC trial, N Engl J Med 2015; aspirin continuation safety.

Disclaimer: This article reflects current evidence but does not replace clinical judgment. Always individualize management per patient factors and institutional protocols.